Systems and methods for signaling neural network post-filter tensor information in video coding

Abstract

A device may be configured to perform filtering based on information included in a neural network post-filter characteristics message. In one example, the neural network post-filter characteristics message includes a syntax element specifying the bit depth of luma sample values output by the post-processing filter and a syntax element specifying the bit depth of chroma sample values output by the post-processing filter.

What is claimed is: 
     
       1. A method of performing neural network filtering for video data, the method comprising:
 receiving a neural network post-filter characteristics message; 
 parsing a first syntax element in the neural network post-filter characteristics message, wherein the first syntax element equal to 0 indicates that sample values output by a post-processing filter are real numbers and the first syntax element equal to 1 indicates that luma sample values and chroma sample values output by the post-processing filter are unsigned integer numbers; 
 parsing a second syntax element in the neural network post-filter characteristics message, wherein the second syntax element plus eight specifies a bit depth of luma sample values in an output integer tensor; and 
 parsing a third syntax element in the neural network post-filter characteristics message, wherein the third syntax element plus eight specifies the bit depth of chroma sample values in the output integer tensor. 
 
     
     
       2. A device comprising one or more processors configured to:
 receive a neural network post-filter characteristics message; 
 parse a first syntax element in the neural network post-filter characteristics message, wherein the first syntax element equal to 0 indicates that sample values output by a post-processing filter are real numbers and the first syntax element equal to 1 indicates that luma sample values and chroma sample values output by the post-processing filter are unsigned integer numbers; 
 parse a second syntax element in the neural network post-filter characteristics message, wherein the second syntax element plus eight specifies a bit depth of luma sample values in an output integer tensor; and 
 parse a third syntax element in the neural network post-filter characteristics message, wherein the third syntax element plus eight specifies the bit depth of chroma sample values in the output integer tensor. 
 
     
     
       3. The device of  claim 2 , wherein the device includes a video decoder. 
     
     
       4. A device comprising one or more processors configured to:
 signal a neural network post-filter characteristics message, wherein the neural network post-filter characteristics message includes: 
 a first syntax element, wherein the first syntax element equal to 0 indicates that sample values output by a post-processing filter are real numbers and the first syntax element equal to 1 indicates that luma sample values and chroma sample values output by the post-processing filter are unsigned integer numbers, 
 a second syntax element, wherein the second syntax element plus eight specifies a bit depth of luma sample values in an output integer tensor, and 
 a third syntax element, wherein the third syntax element plus eight specifies the bit depth of chroma sample values in the output integer tensor. 
 
     
     
       5. The device of  claim 4 , wherein the device includes a video encoder.